Use of the Delphi method as an instrument of community participation in health needs assessment

Vol:.(1234567890)
Jou nal o Public Heal h Policy (2025) 46:460–472
h ps://doi.o g/10.1057/s41271-025-00559-9
VIEWPOINT
Use o  heDelphi me hod asanins umen o communi y
pa icipa ion inheal h needs assessmen
LunaRod íguezPé ez1· ManuelRod íguezRod íguez2·
Rosa ioVigoO ega3 · Sil iaSic eAlonso4· T ánsi oCeb iánVale o5·
Pila Men uyIsus5· LuisGab ielLuqueRome o6,7,8
Accep ed: 16 Janua y 2025 / Published online: 29 Ap il 2025
© The Au ho (s) 2025
Abs ac
We conduc ed a comp ehensi e analysis o he use ulness o he Delphi echnique
o acili a ing communi y pa icipa ion in local heal h needs assessmen s wi hin he
Andalusian Local Heal h Ac ion Ne wo k, Spain. We de eloped an ad hoc online
ques ionnai e based on he Social De e minan s o Heal h model and applied i o a
panel o expe s in wo municipali ies in he p o ince o Se ille (Andalusia, Spain)
be ween May and June 2021. Ou esul s e lec ed good panelis pa icipa ion. The
ques ionnai e success ully enabled he p io i iza ion o bo h new and o iginal i ems,
some o which we e inco po a ed in o local heal h policies. We concluded ha
he Delphi me hod was e ec i e o acili a ing pa icipa ion in local heal h needs
assessmen s o e ing a eplicable, cos -e ec i e app oach ha accele a ed local pol-
icy de elopmen and suppo ed he implemen a ion o Heal h in All Policies wi hin
local go e nmen .
Keywo ds Heal h policy· Social de e minan s o heal h· Delphi echnique· Social
pa icipa ion· Communi y pa icipa ion
Key messages
• The Delphi Me hod (DM) is a cos -e ec i e me hod o assessing and p io i iz-
ing heal h needs wi h communi y pa icipa ion.
• The DM helps o de ine local heal h policies in dispe sed o unde - esou ced
a eas.
• A Social De e minan s ques ionnai e acili a es local implemen a ion o Heal h
in All Policies.
Ex ended au ho in o ma ion a ailable on he las page o he a icle
461
Use o  heDelphi me hod asanins umen o communi y…
In oduc ion
Since he elease o he O awa Cha e o Heal h P omo ion in 1986, subs an-
ial e idence has eme ged highligh ing he c i ical ole o social de e minan s
in in luencing heal h ou comes [1, 2]. As a esul , in e sec o al ac ion has been
inc easingly p omo ed o imp o e he quali y o li e o he gene al popula ion [3].
In e na ional [4], Eu opean [5], and Spanish [6] ins i u ions ha e endo sed he
Heal h in All Policies (HiAP) app oach [4], which equi es policymake s in all
sec o s o sys ema ically conside he heal h impac s o hei decisions [7]. Local
go e nmen s play a c ucial ole in his amewo k, as hey a e he le el o go e n-
ance closes o he popula ion and can implemen policies ha mo e accu a ely
e lec local needs [8]. The de elopmen o HiAP a he local le el is o en based
on heal h needs assessmen s (HNA), which ely on communi y pa icipa ion [8]
o ensu e ha policies esona e wi h ci izens [10] and o design in e en ions
aimed a imp o ing heal h [9].
In ou analysis, we explo ed he use o he Delphi me hod o acili a e com-
muni y in ol emen in he HNA p ocess. This ini ia i e was conduc ed wi hin
he amewo k o he Local Heal h Ac ion Ne wo k (RELAS), a p ojec spon-
so ed by he Andalusian Minis y o Heal h and Consume A ai s o p omo e he
implemen a ion o HiAP by de eloping local heal h plans (LHPs) based on heal h
needs assessmen da a [9], consis ing o bo h quan i a i e da a (demog aphic,
epidemiological, economic, geog aphical, and social indica o s) and quali a i e
insigh s ga he ed om he communi y h ough pa icipa o y me hods such as
ocus g oups, su eys, and in e iews. A s ee ing g oup hen analyses his in o -
ma ion o p io i ize heal h needs and guide he de elopmen o in e en ions [9].
In he RELAS amewo k, his s ee ing g oup, known as he Technical Ad i-
so y G oup [11], comp ises membe s om a ious heal h adminis a ion a eas,
wi h leade ship p o ided by a p o incial-le el Expe Technical Ad iso in
heal h p omo ion. The Ad iso suppo s and guides municipali ies h oughou he
p ocess.
Du ing he COVID-19 pandemic in sp ing 2021, adi ional pa icipa o y ools
we e imp ac ical due o he need o ace- o- ace in e ac ion. To p e en dis up-
ion o he p ocess and ensu e communi y inpu , he Heal h P omo ion Team o
he Alja a e-No h Se ille Heal h Dis ic —composed o he Expe Technical
Ad iso , he Di ec o o Disease P e en ion, Heal h P omo ion and Su eillance
Uni , and leade s om he Local Heal h Ac ion Ne wo k—op ed o use he Del-
phi me hod.
We sough o analyze he implemen a ion p ocess and esul s o ou expe ience
wi h he Delphi me hodology o p omo e in e sec o al pa icipa ion in he de el-
opmen o local heal h policies in wo municipali ies in he p o ince o Se ille,
Andalusia.
462
L.R.Pé ez e al.
Da a andme hods
The Delphi echnique is widely used in heal h sciences esea ch o iden i y p i-
o i ies and each g oup consensus on key issues [12]. I is a s uc u ed me hod
designed o ga he expe opinions on a gi en opic [13]. In ou analysis, we ep-
lica ed key ea u es o he Delphi me hod: pu posi e sampling, pa icipan ano-
nymi y, i e a ion, and con olled eedback [14].
Fieldwo k was conduc ed be ween May and June 2021 in wo municipali ies
wi hin he Alja a e-No h Se ille Heal h Dis ic (Mai ena del Alco and Mai ena
del Alja a e). A e ailing o ind an exis ing ins umen sui ed o ou objec i es,
we designed an ad hoc online ques ionnai e based on Dahlg en and Whi ehead’s
(1991) amewo k o heal h de e minan s [14]. This amewo k is widely used
due o i s simplici y and p ac icali y. The ques ionnai e, e iewed by expe s and
public ep esen a i es, was app o ed wi h only mino e isions.
The inal e sion included 48 closed ques ions ac oss six componen s: habi s
and li es yle; educa ion sys em; social en i onmen ; socioeconomic condi ions;
physical en i onmen and neighbo hood; and heal hca e sys em. Pa icipan s
a ed he impo ance o each i em on a 5-poin Like scale ( om 0 = leas impo -
an , o 5 = mos impo an ) and had he op ion o sugges new i ems ele an o
hei municipali y.
In Rounds 1 and 2, pa icipan s a ed he same i ems, along wi h new ones sug-
ges ed a e Round 1. The impo ance a ings om Round 1 we e p o ided o
e e ence, and he same c i e ia we e used o e- a ing. Each ound las ed wo
weeks, wi h email eminde s sen o non- esponden s o ensu e pa icipa ion.
The Expe Technical Ad iso p ocessed he da a and p epa ed a epo on he
esul s. The p io i ies iden i ied h ough he Delphi p ocess we e added o he
HNA and included in he local heal h plan, adop ed by local councils o suppo
he HiAP s a egy.
The Heal h P omo ion Team de ined he selec ion c i e ion o expe s as
membe ship o he communi y, conside ing indi iduals who li ed o wo ked in
he municipali y (geog aphical c i e ion) o belonged o a local sec o o g oup
(sec o c i e ion). The selec ed expe p o iles included ep esen a i es om ci i-
zens’ associa ions, key in o man s amilia wi h he municipali y ( o example,
ep esen a i es o poli ical g oups, ade unions, and employe s’ associa ions),
municipal echnical s a ( om depa men s such as housing, en i onmen , and
public sa e y), s a om public ins i u ions (schools, social se ices), p ima y
ca e p o essionals (doc o s, nu ses, social wo ke s), communi y pha macis s, and
ep esen a i es o ulne able g oups.
The Heal h P omo ion Team con ac ed po en ial pa icipan s by phone o in i e
hem, explain he pu pose o he analysis, and in oduce he Delphi me hodol-
ogy. Indi iduals who demons a ed willingness, commi men , ime a ailabili y,
and communica ion skills [15] we e included. Those who ag eed o pa icipa e
ecei ed an email om he Expe Technical Ad iso wi h ins uc ions, a link o
he ques ionnai e, and he esponse deadline.
463
Use o  heDelphi me hod asanins umen o communi y…
Da a analysis
We de ined consensus in ad ance [16] and es ablished wo c i e ia: consis ency
o esponses be ween ounds and he absence o new i em con ibu ions. To check
consis ency, we applied he K uskal–Wallis es , compa ing median sco es o
each i em ac oss ounds, wi h a signi icance le el o p ≤ 0.05. To p io i ize i ems,
we used wo condi ions: a median sco e o 4 o 5 (indica ing high impo ance)
and an accep able consensus le el ( ela i e in e qua ile ange (RIR) ≤ 75%) [17].
Da a om he online ques ionnai e we e ans e ed o Open O ice sp eadshee s,
and s a is ical analyses we e pe o med using Epi In o 7 so wa e [18].
Resul s
Since ou main objec i e was o analyze he use ulness o he Delphi me hod o
acili a ing communi y pa icipa ion in local HNAs, we emphasize he applica-
ion o he me hodology. We also p o ide a b ie analysis o he p io i ized i ems
as hey ela e o acili a ing he de elopmen o local heal h policy.
The e was a high le el o pa icipa ion ela i e o he ini ial numbe o people
in i ed. The a i ion a e be ween ounds was highe among male pa icipan s
(Table1). The dis ibu ion o i ems pe heal h de e minan a ies by municipali y,
e lec ing he addi ion o new i ems ha co espond o speci ic local cha ac e is-
ics (Table2). The dis ibu ion o p io i ized i ems co esponds o he pe cen age
o hose included in he o iginal ques ionnai e and hose p oposed by panelis s
(Table3).
The s ee ing g oup included some o he p io i ized i ems (4 in Mai ena del
Alco , 5 in Mai ena del Alja a e) in he LHPs, which he municipal councils hen
app o ed. Th ee p io i ized p oblems associa ed wi h habi s and li es yle (poo
die , seden a y li es yle, s ess) we e common o bo h municipali ies. In he case
o Mai ena del Alja a e, p oblems ela ed o he heal h sys em we e excluded
om he LHPs and we e e e ed o he managemen eam o he e e ence Heal h
Dis ic .
Table 1 Pa icipan lows by sex and municipali y
Mai ena del Alco Mai ena del Alja a e
Round 1 Round 2 Round 1 Round 2
Numbe in i ed 46 (100%) 35 (100%) 45 (100%) 28 (100%)
Numbe o pa icipan s 35 (76.0%) 25 (71.4%) 28 (62.2%) 26 (92.8%)
Sex
Female 31 (68.0%) 18 (70.0%) 26 (57.8%) 19 (73.3%)
Male 15 (32.0%) 7 (30.0%) 19 (42.2%) 7 (26.7%)
464
L.R.Pé ez e al.
Table 2 E olu ion o i ems add essed in wo Delphi ounds
Mai ena del Alco Mai ena del Alja a e
I ems g ouped by SDH
heme
Included in
ques ion-
nai e
P oposed
by pan-
elis s
No. p io i ized o hose
included in he ques-
ionnai e
No. p io i ized
o hose p o-
posed
Included in
ques ion-
nai e
P oposed
by pan-
elis s
No. p io i ized o hose
included in he ques-
ionnai e
No. p io i ized o
hose p oposed
Habi s and li es yles 7 2 3 0 7 4 5 1
Educa ion sys em 9 8 0 0 9 6 0 0
Social en i onmen 7 2 1 2 7 6 0 0
Socioeconomic condi-
ions
6 3 5 1 6 3 6 1
Physical en i onmen
and neighbo hood
13 3 0 0 13 15 0 3
Heal h sys em 6 3 2 2 6 6 2 6
To al 48 21 11 5 48 40 13 11

465
Use o  heDelphi me hod asanins umen o communi y…
Table 3 I ems included and p io i ized a e wo Delphi ounds
*Only p io i ized in Round 2
I ems Round 1 I ems Round 2
No. included in
ques ionnai e
No. p oposed by
panelis s
No. included in
ques ionnai e
No. p oposed by
panelis s
No. p io i ized o hose included
in he ques ionnai e
No. p io i ized
o hose p o-
posed
Mai ena del Alco 48 21 69 0 11* (22.9%) 5* (23.8%)
Mai ena del Alja a e 48 40 88 0 13* (27.1%) 11* (27.5%)
466
L.R.Pé ez e al.
Discussion
Ou analysis aimed o examine he e ec i eness o he Delphi me hod in p o-
mo ing pa icipa ion in Heal h Needs Assessmen s a he local le el. I add esses
a gap in he li e a u e conce ning he lack o conc e e expe iences wi h Heal h
in All Policies, especially ega ding implemen a ion and p ac ical ecommenda-
ions [8, 19, 20], which limi s local policy de elopmen wi hin his pa adigm [8].
The li e a u e also emphasizes he need o esea ch on acili a o s and ba ie s o
HiAP s a egy de elopmen [8], pa icula ly in he pa icipa ion phase [19], a c u-
cial elemen ha in luences policy decisions in he HiAP p ocess [8, 20]. While
ocus g oups and key in o man in e iews a e commonly epo ed echniques,
he use o he Delphi me hod o HNA is ela i ely new [9].
We ound ha he Delphi me hodology was an e icien way o ob ain a consen-
sus lis o SDHs in bo h municipali ies. The li e a u e suppo s he ad an ages o
ga he ing in o ma ion h ough his in e ne -based ool compa ed o ace- o- ace
pa icipa o y echniques, highligh ing i s low cos and sa ings on ypical expenses
like pa icipan a el and enue hi e [21, 22]. Ra aghi e al. (2023) iden i ied
se e al challenges in conduc ing an HNA ha Delphi can add ess [9]. Fi s , Del-
phi is eplicable and can engage he communi y wi hou sac i icing me hodologi-
cal alidi y [23]. Second, he use o new echnologies educes logis ical ime and
sa es human and economic esou ces [16, 24]. We iden i ied a po en ial ba ie in
he HNA p ocess: local heal h p o essionals may lack skills in building consensus
h ough pa icipa o y decision-making. To mi iga e his, an ex e nal agen could
adminis e he ques ionnai e o pa icipan s. Some au ho s ha e no ed challenges
in ensu ing adequa e pa icipa ion and he isk o legi imizing p e-made decisions
wi hou allowing di e se ci izen iews o eme ge [10]. The Delphi me hod can
esol e his by acili a ing pa icipa o y dialog o e se e al ounds and allowing
panelis s o add hei own opinions h ough open-ended ques ions.
In he RELAS p ojec , we iden i ied he ollowing ad an ages. Fi s , he Del-
phi me hod accele a es he pa icipa ion phase compa ed o o he me hods, when
conduc ing an HNA akes so long ha i delays he p epa a ion o he LHP [23].
Using his me hod, we educed he o e all p ocess om an a e age o h ee yea s
o one yea . Second, wi h an ex e nal echnical ad iso acili a ing he p ocess, a
single pe son can apply he ool in mul iple communi ies. Thi d, i is possible o
inc ease he numbe o panelis s wi hou a co esponding inc ease in complexi y.
Finally, he eplicabili y o he echnique allows us o iden i y changing needs
wi hin he same communi y and o compa e di e en communi ies while iden i y-
ing common ac o s in luencing local needs. This me hod could play an impo -
an ole in heal h a eas whe e communi ies a e sca e ed, whe e i is di icul o
key in o man s o a el, o whe e he cos o human and economic esou ces is
p ohibi i e.
We belie e ha communi y HNAs should ocus on he social de e minan s o
heal h [9]. Acco dingly, we designed ou ini ial ques ionnai e a ound Dahlg en
and Whi ehead’s model, which has been success ully used by p ac i ione s and
policymake s ou side he heal h sec o o concep ualize de e minan s [14]. Ou
467
Use o  heDelphi me hod asanins umen o communi y…
ques ionnai e included a speci ic lis o i ems o each de e minan o encou age
panelis s o conside he b oade con ex o social de e minan s o heal h when
esponding and o s imula e he sugges ion o new i ems o each de e minan .
While he leng h o he ques ionnai e may con ibu e o pa icipan a i ion [25],
he e a e no explici c i e ia o de ining an op imal numbe o i ems o p e en
his. Following Lilly e  al.’s ecommenda ion, we g ouped i ems in o blocks
[21]—in ou case, by ype o de e minan — o help panelis s main ain a de e -
minan s-based pe spec i e o local eali y and p opose new i ems o each block
[14].
We acknowledge ha he c i e ion o inclusion as an expe panelis is con-
o e sial in he li e a u e due o he challenge o measu ing expe ise. The e o e,
i is necessa y o de ine his c i e ion be o e o ming he panel [13, 26]. Follow-
ing his ecommenda ion, we es ablished communi y membe ship [9] as ou c i-
e ion, using geog aphy as he p ima y e e ence and ep esen a ion o a ious
sec o s and local g oups (heal h, educa ion, he elde ly, child en, equali y, e c.)
as seconda y. We conside ed panel membe s o be expe s because hey li e o
wo k in he municipali ies and possess i s -hand knowledge o he local eali-
ies. In he scien i ic li e a u e, such indi iduals a e e e ed o as a ec ed pa ies,
in ol ed pa ies, o key in o man s in diagnos ic p ocesses who ha e knowledge
o he si ua ion unde analysis ega dless o i le, s a us, o ank. They con as
wi h expe s, who mus ha e ou s anding quali ica ions in he subjec unde s udy,
such as academic c eden ials, special quali ica ions, o no able p o essional expe-
ience [27]. We se an ini ial minimum o 40 panelis s, an icipa ing a i ion in
subsequen ounds, and hus ensu ed a di e se ange o pa icipan iews in bo h
municipali ies. While he e is no consensus on he op imal numbe o panelis s,
be ween 10 and 30 is gene ally conside ed su icien [13, 20, 26].
A e analyzing he p io i iza ion esul s, we no ed ha he same h ee i ems
ela ed o li es yle habi s we e p io i ized and included in he LHPs o bo h
municipali ies [26]. This may be because hese de e minan s a e ob ious and well
known, sugges ing he e is s ill oom o imp o e he dissemina ion o he Heal h
in All Policies app oach. Con ibu ions made du ing he Delphi p ocess could
help he s ee ing g oup iden i y new local eali ies equi ing in e en ion.
The li e a u e sugges s ha ansla ing Heal h Needs Assessmen indings in o
speci ic policies has been e y limi ed. Ra aghi e al. (2023) conduc ed a scoping
e iew o 169 s udies: only i e men ioned success ul collabo a ion be ween he
heal h sec o and policymake s, and only ou succeeded in se ing o changing
policy agendas [9], highligh ing he lack o linkages be ween p oposals esul -
ing om pa icipa ion and heal h policies [10]. The li e a u e also men ions ha
poli ical ep esen a i es o en do no us he esul s o ci izen pa icipa ion
ocused on social de e minan s o heal h [28]. In hei 2023 e iew, Ling e al.
discussed ac o s in luencing he HiAP p ocess in local go e nmen s and no e
he ambigui y and di e si y o aspec s a ec ing poli ical decision-making, which
may be one eason why some issues p io i ized by ci izens a e no included in
LHPs [28]. This was he case wi h i ems ela ed o heal h se ices in Mai ena del
Alja a e, which we e excluded en i ely.
468
L.R.Pé ez e al.
S udy limi a ions
We iden i ied mos o he limi a ions in ou analysis as ela ing o he panelis s. Fi s ,
because we did no s ipula e a minimum numbe o ep esen a i es pe sec o bu
a he a o al numbe , some sec o s may ha e been o e - o unde ep esen ed, po en-
ially a ec ing he numbe o p io i ized i ems pe de e minan (mo e heal h sec o
panelis s p io i izing mo e heal h- ela ed issues). Second, we we e unable o e i y
he le el o expe ise o he panelis s. Since i is di icul o quan i y communi y
membe ship, we could no de e mine whe he indi iduals li ing o wo king in he
communi y we e knowledgeable abou i , especially i hey had ecen ly mo ed he e
[16]. Thi d, ensu ing he independence o he panelis s was challenging, as pa ici-
pa ion was limi ed o hose wi h a p e-exis ing in e es in he opic o a ela ionship
wi h he p ojec [25]. This ela es o he di icul y o con olling o bias om pan-
elis s wi h pe sonal o sec o al iews o local eali ies [9]. Fou h, we did no di ec ly
check ha he panelis s unde s ood he i ems included in he ques ionnai es. Di e -
ences in in e p e a ion may limi consensus using his echnique [16]. Conduc ing an
online session ia ideocon e ence be o ehand o explain he p ojec and p ocedu es
could mi iga e his limi a ion. Fi h, al hough we included local s akeholde s, lead-
e s, and key igu es wi h a b oad ision o a ious needs and social g oups, we did
no conside including indi iduals who lacked In e ne access, had limi ed li e acy,
o belonged o pa icula ly impo an disad an aged g oups in he locali y. Fu he -
mo e, he ini ial alida ion o he ques ionnai e was a e y simpli ied p ocess due
o he lack o a simila , eady-made ool. As a esul , he ques ionnai e may ha e
included i ems ha we e oo gene al o open o di e en in e p e a ions. We do no
know whe he he lis o i ems p io i ized by he panels would ha e been di e en i
o he echniques had been employed. Fu he mo e, we we e unable o compa e ou
esul s, e en pa ially by sec o , because he e a e no o he simila local publica ions
on hese opics.
A common limi a ion o pa icipa o y p ocesses is he e hical challenge o man-
aging pa icipan s’ expec a ions abou he ul ima e impac o hei con ibu ions
[9]. Al hough we made i clea h oughou ha Delphi p io i iza ion was pa o
a b oade p ocess, we did no e alua e panelis s’ pe cep ions o he p ocess o he
esul s.
Fu u e di ec ions
Wi h his analysis, we ha e pa ed he way o u he de elopmen in designing and
alida ing an ini ial ques ionnai e model ha can be used o conduc Heal h Needs
Assessmen s wi h a ocus on Social De e minan s o Heal h in local se ings. Based
on a simple, conc e e, and acc edi ed me hodology, his ool would p o ide local
go e nmen s wi h he oppo uni y o implemen Heal h in All Policies and o assess
whe he channeling ci izen pa icipa ion in he HNA p ocess h ough a speci ic ali-
da ed ool enables policymake s o eel mo e com o able implemen ing HiAP a he
local le el.